Designing High Speed Backplanes
Utilizing Physical Layer Test System
and Advanced Design System Tools
Application Note
Abstract
Telecommunication systems are required to achieve high-speed data transport
up to data rates beyond 10 Gb/s. One of the biggest challenges facing digital
designers today in their quest for 10 Gb/s and beyond is that of backplane
packaging. An emerging process in the design cycle for the high-speed digital
designer includes the use of measurement-based modeling. The success level
of this methodology can be very dependent upon which tools are employed
in the design lab and how well they work together. This paper will look at
the problems introduced into the backplane assembly design by the many
linear passive components that create reflections due to impedance discontinuities. Using a typical backplane design case study utilizing two popular
design tools, the Physical Layer Test System (PLTS) and the Advanced
Design System (ADS), we will evaluate the benefits of the measure, model,
and simulate methodology and the advantages it provides to the practical
designer.
2
Typical 10 Gb/s Telecom System
Figure 1. Typical high-speed networking applications.
Today’s advanced telecommunication systems are required to achieve
high speed data transport up to data rates beyond 10 Gb/s. Modular chassisbased systems, such as core routers, Ethernet switches, and storage subsystems are becoming evermore challenging for the digital design engineer
due to signal integrity problems in the physical layer. Printed circuit boards,
connectors, cables, and vias used to be simple components that were easily
incorporated into a system with little or no thought. Now, the data risetime
from a zero logical level to a one logical level can be less than 100 ps, thus
creating microwave transmission line effects in formerly simple structures.
In most high-speed networking applications such as those shown in Figure 1,
the chassis-based system interface will be used to carry communications
between the control-plane processors on each line card. This physical layer
copper interface creates many challenges for signal integrity engineers
designing, developing, and testing network elements. One of the most
challenging and interesting areas for high speed design is in backplane
applications. Performance of routers and switches are fundamentally limited
by the bottleneck created by the backplane components and therefore this is
an area rich for technology breakthroughs and innovation.
An emerging process in the design cycle for the high speed digital designer
includes the use of measurement-based modeling. Many design tools are
available on the market today that support the iterative process of measure,
model, and simulate. The success level of this methodology can be very
dependent upon which tools are employed in the design lab and how well
they work together. This paper will discuss a typical backplane design case
study utilizing two popular design tools, the Physical Layer Test System
(PLTS) and the Advanced Design System (ADS).
3
Impedance Problems are Everywhere!
Figure 2. Examples of telecom system passive components.
The physical layer within a telecom system has many linear passive components that create reflections due to impedance discontinuities. As shown
in Figure 2, these systems consist of backplane assemblies that contain
redundant switch cards, line cards, connectors, IC packages, printed circuit
boards and power supply modules. The copper backplane itself is probably
the most challenging design project due to the fact that it must have the
longest lifecycle in the field. A total router system upgrade allowing a
complete retrofit of a backplane is unlikely, therefore the most advanced
backplane design tools and methods are required to give the highest backplane performance so as to achieve an extended lifetime in the field. All
modular chassis-based systems feature a high-speed backplane and multiple
line cards. Performance and capacity improvements are done by adding
more line cards and increasing the port density on the line cards. These
systems are modular and scale independently without requiring an entire
forklift upgrade or replacement. These telecom systems must be designed
for high availability with redundant components to ensure uptime.
4
Agilent’s Signal Integrity Portfolio
Figure 3. Relationship of Agilent test equipment and software tools.
There are many ways to design high speed digital networks and many tools
to accomplish this task in the laboratory. As a general overview, Figure 3
describes the Agilent test equipment and software tools that are used today
in many phases of advanced design, development, debug, validation, and
standards compliance testing. The area that we will focus on with this
application note is the physical layer. More specifically, we will go into detail
about the passive interconnect measurements, device characterization, and
modeling. A strong synergy can be developed with specific tools within the
design cycle that enables a measurement-based modeling approach. The
PLTS and ADS tools have cross functional compatibility with importing
and exporting data.
5
Where Today’s Solutions Are Used
Figure 4. Three phases of design.
The methodology of measure, model, and simulate that has been developed,
provides distinct advantages for the practical designer. The iterative process
of the three key phases of design shown in Figure 4 allow refinement and
optimization, ultimately reducing the overall backplane design cycle time.
The value of design software is to reduce the number of iterations through
the build and test methodology by providing an environment for virtual
testing. The time required to design, produce, and measure is long, and the
associated expenses are large. However, there are many practical advantages
to producing intermediate prototypes, such as simplifying complex electrical
interactions into a measurement-based model, and providing feedback to
verify the designer’s knowledge of the important physical effects.
Once the first prototype is fabricated, then various instrumentation can
be used to obtain a measurement-based model, which then can be easily
imported directly into common simulators. PLTS, which controls and
calibrates either a Vector Network Analyzer (VNA) or a Time Domain
Reflectometer (TDR), will make the measurement and then export the
appropriate format of data to the simulator application, ADS software.
Touchstone or Citifile are the most common PLTS exported file types.
6
Simulation Engines for the SI Designer
Frequency
Advanced Design System
AC/Sparameters
Vector signal
instrument
links
Harmonic
balance
Physical
Numeric
Time
Circuit
envelope
HF
SPICE
(transient)
Convolution
Ptolemy
synchronous
dataflow
Circuit and numeric
co-simulation
High speed
interconnect
library
Momentum
(3D planar)
Layout
components
Ptolemy
timed
synchronous
data flow
(TSDF)
Advanced
model
composer
3D EM
simulator
Capabilities
Figure 5. ADS environment.
ADS is a technology rich simulation environment. The most important
simulation technologies for signal integrity work are highlighted in Figure 5,
with Ptolemy being given specific emphasis. Ptolemy is used to simulate the
complete transmitter to receiver link, including coding, pre-emphasis, and
equalization techniques. Circuit level co-simulation allows the inclusion of
linear and non-linear circuit based models and data based models, including
IBIS. Co-simulation at the numeric level allows the inclusion of Matlab® and
VHDL models. This co-simulation is a key component to simulating the
complete differential backplane channel. Also, physical simulation is tightly
integrated into the ADS environment with electromagnetic analysis integrated
into the layout and schematic environment.
7
Backplane Design Flow
Figure 6. Typical design process.
The design flow for backplanes shown in Figure 6 starts by constructing
abstract blocks to define the overall behavior of the system, and then moves
to more detailed designs as specific circuits become available. The backplane
design process starts by knowing the system performance requirements for
the standard being implement. System requirements will determine the
need for pre-emphasis or equalization, typical or maximum distances for
transmitting a signal, and what type of coding/decoding is included in the
signal. Ptolemy can be used to simulate the entire link from transmitter to
receiver, including coding and pre-emphasis and equalization techniques.
Ptolemy is described as the Integration simulator, with it’s ability to cosimulate with analog models from ADS, with Matlab models, with Verilog-A
models, and with RTL-HDL. Designers can start with intellectual property
(IP) from a variety of these sources, and integrate them into a single
simulation using Ptolemy.
The next step is to partition the system design to the specifications for
individual blocks. The designer starts to move from the more abstract specs
toward circuit-level implementation. The typical process might look like the
following;
1. The system designer partitions the design and gives the specifications
to the circuit designer for input signals, power consumption, and output
signal from the circuit in terms of signal level, noise and allowable
error rate.
2. The initial circuit design starts with ideal signal sources with impedance
defined for the source and load, and with ideal connections between
components.
3. Then, as the basic functionality of the analog circuit is determined,
some of the connections are replaced with more realistic transmission
lines as knowledge is gained about the actual connection between
components (using microstrip / stripline).
4. Finally, replace individual transmission lines with coupled lines that
interact (added parasitic effects, coupling, impedance discontinuities).
8
One thing that enables complete system design is a powerful capability
called co-simulation. Co-simulation is a way to combine models that are best
simulated in one type of simulator with the models from another type of
simulator. For example, system behavioral models are best simulated with a
numeric domain simulator such as ptolemy, but microstrip transmission
lines are best described with frequency domain analog simulators, such as
Linear or Harmonic Balance. ADS co-simulation enables a single schematic
to contain system behavioral models along with detailed analog circuit-level
models. ADS enables co-simulation between system (Ptolemy), analog circuit
(Transient, Convolution, Circuit Envelope), and EM physical simulators
(Momentum). Behavioral models occur both in system simulation as well as
analog simulation.
In terms of analog circuit simulation for high-speed backplane design, the
most commonly considered simulators are the time-domain simulators
(Transient and Convolution). Other simulators, such as the frequency domain
S-Parameter and Harmonic Balance simulators, can be used for some
aspects of the backplane depending on the nature of the test. For example,
Harmonic Balance is used for fast simulations if there is a steady-state
periodic signal which includes active devices. S-parameter analysis is used
for determining the frequency response of the transmission path. Physical
simulation is tightly integrated into the ADS environment with electromagnetic
analysis integrated into the layout and schematic environment.
Throughout the design cycle, it is helpful to construct smaller circuits to test
your understanding of the important electrical effects. Smaller test circuits
can be used as quick verification of simulations and they can be used as
measurement-based models within the simulator. ADS has many ways to
link physical tests back into the simulator, including S-Parameters from
PLTS as mentioned, or capturing waveforms from scopes, such as the DCA-J
(86100C). Eventually, after enough design knowledge has been built up
through the iterative use of circuit simulation, electromagnetic analysis, and
measurement-based modeling, the designer should return to the system
simulation. It is within this environment that the designer should update the
system design to include the more specific designs created (check to see if
the system performance is maintained per the specification). If compliance
to the specifications is achieved, then the backplane is ready for the final
build and test. This may mean that the layout produced by ADS is the final
layout ready for building, or it may mean linking to other tools for layout
through formats such as Gerber or GDS-II.
9
Physical Layer Test System
Figure 7. The Physical Layer Test System.
Physical Layer Test System has been designed specifically for signal integrity
analysis. As shown in Figure 7, PLTS software guides the user through
hardware setup, calibration, and data acquisition. Both Time Domain
Reflectometers or Vector Network Analyzers can be used as the measurement
engine and the calibration wizard for each will allow advanced calibration
techniques. This is helpful when trying to remove unwanted test fixture
effects such as cable loss, connector discontinuities, and dielectric loss of
printed circuit board material. The PLTS device data base is used to view
the performance characteristics in many useful ways. Frequency domain
analysis is now becoming mandatory for high speed digital standards such
as PCI Express II and Serial ATA. Since high speed data has fast risetime
edges that create microwave transmission line effects within the backplane
channel, input differential insertion loss (SDD21) is now a required test for
compliance. The familiar time domain analysis (TDR &TDT) is complemented
by a Novel eye diagram synthesis engine. A virtual bit pattern generator feature
allows either a user-defined binary sequence or standard PRBS to be applied
to the measured data to convolve eye pattern diagrams. In addition, PLTS
applies patented transformation algorithms to present the data in both
frequency and time domains, forward and reverse directions of signal flow,
transmission and reflection terms, and in all possible modes of operation
(single-ended,differential,and mode-conversion).
10
Measurements on Various Trace Lengths
Figure 8. Measured differential insertion loss.
Once the hardware components in the complete backplane channel are
locked down, a prototype can be built and evaluated. In the measurements
shown in Figure 8, a press fit connector was used to build up a backplane
test vehicle. The PLTS system was utilized to measure all 4-port parameters
and the differential insertion loss (SDD21) for three different length traces.
Most digital standards are now requiring this SDD21 as a performance
figure-of-merit. This parameter can be thought of as the frequency response
seen by the differential signal as it propagates through the backplane channel.
The amount of attenuation verses frequency is a good way to judge performance. Naturally, the shorter the channel, the less the attenuation as a function
of frequency.
11
Mode Conversion
Figure 9. TCD11 to TDD11 aligned in time.
One of the biggest challenges that delays time to market for backplane
designs is the presence of crosstalk within the channel (between one differential pair to an adjacent differential pair). Of course, crosstalk within a pair
is highly desirable and we call that coupling. This strong coupling provides
high common mode rejection ratio (CMRR). However, pair to pair, any
mode conversion will create crosstalk. A practical application of how mode
conversion can be identified in physical layer devices is shown in Figure 9.
This shows a XAUI backplane with two daughter cards that typically transmit
data at 3.125 Gb/s. The design objective for this high-speed differential
channel is to minimize the crosstalk between adjacent differential PCB traces
throughout the length of the channel.The channel consists of the linear
passive combination of the backplane and two daughter cards. Any mode
conversion from differential mode to common mode will generate EMI and
create crosstalk that will be incident upon other channels and will degrade
performance. Locating the exact structure within the channel that creates
the most mode conversion is a useful debug tool for backplane design.
Looking at Figure 9, the differential to common mode conversion time
domain reflection parameter (TCD11) is time aligned with the differential
impedance profile of the channel (TDD11) below it. A marker is placed on
the largest magnitude peak of TCD11.This is where the physical structure
within the channel is creating the most mode conversion and thus the
source of the most crosstalk.We can align the TDD11 to the TCD11 in time
and therefore spatially co-locate the problematic structure on TDD11.To
relate this structure to the channel, we use the differential impedance profile
as a reference. It is known that the two capacitive discontinuities on TDD11
are the daughter card via field and motherboard via field, respectively. Since
the marker falls upon the second discontinuity on TDD11, it is discovered
that the motherboard via field is the biggest culprit to causing crosstalk in
adjacent channels. The motherboard via field should be re-routed and the
crosstalk generation will be reduced. This measurement and analysis can be
done completely within the PLTS tool environment.
12
Backplanes are a Critical Link
Figure 10. Examples of typical backplanes.
The telecommunications network design engineer does not usually have the
flexibility to design the complete system from the ground up. A more realistic
situation is where a portion of the system is a legacy, that is, some network
components exist and must be utilized in the new design. For example, a
new transmitter line card must be developed for a legacy backplane within a
ethernet router. The receiver card has already been designed and all that is
required is a new transmitter card. In this particular case, the designer
would like to simulate how his newly developed transmitter will function
within the existing system. Rather than developing a complex model of the
backplane channel that consists of complex via structures and connectors,
a measurement can be made with PLTS. The 4-port s-parameter data from
the measurement can be exported from PLTS as a Touchstone file and
subsequently imported into ADS. This saves a tremendous amount of time
and yields a very accurate behavioral model for a complex structure. This
synergy saves modeling time, produces accurate simulations, and gets the
transmitter card to market faster. Figure 10 shows examples of typical
backplanes for various high speed serial protocols, including Extended
Attachment Unit Interface (XAUI) and Advanced Telecommunications
Architecture (ATCA).
13
System Simulation Setup
Figure 11. Comparison of two different backplane/connector configurations.
Some decisions need to be made regarding the backplane architecture;
namely, whether a star, mesh, or other topology is required. Furthermore,
component selections need to be made based on performance cost and
manufacturability requirements. Since the via field in a backplane is usually
the most challenging structure affecting signal integrity, we will spend some
time analyzing this area. In Figure 11, a comparison is made of two different
backplane/connector configurations using either microvias in conjunction
with a SMT connector or through hole vias with a pressfit connector. In a
first step, both alternative scenarios are evaluated in the circuit simulator.
The original goal of this investigation is to demonstrate the influence of
several design parameters:
• distance between daughter cards
• connector to board interface
• width of transmission lines
• PCB material
• metalization layer in the PCB
• termination of the transmission lines
• type of transmission lines
• statistical variations
For this special application, two typical cases will be highlighted which
demonstrate the superior behavior of micro via technology and SMT connectors compared to traditional pressfit connectors and through hole vias. The
following general system scenario in Figure 11 is used for this simulation
study. The signal path is a point-to-point connection between two integrated
circuits on two different daughter cards. The daughter cards are plugged
into a backplane printed circuit board. The connections are done using a
connector with pressfit pins and alternatively a connector with a surface
mount interface to the backplane. The characteristic impedance of differential
striplines is Z0diff = 100 W and the data rate applied is 10 Gb/s using a
Non-Return Zero (NRZ) 8B10B code. The table in Figure 9 lists the different
design parameters of the two experimental test vehicles.
14
System Simulation Setup
Figure 12. Exploded view of SMT backplane connector.
Designing a surface mount backplane connector has numerous challenges.
First, the interface must withstand the mechanical conditions faced by
standard board applications and be very rugged. Secondly, the connector
must be able to transmit data at speeds exceeding 10 Gb/s. Recent designs
of Surface Mount (SMT) backplane connectors have evolved from press fit
connector technology including many of the same mechanical features such
as the 1.5 mm x 2.5 mm pin grid. The two main differences in the connector
designs revolve around the use of SMT signal leads and the ‘C-shaped’
pin-in-paste ground shield pin. A well designed high speed board to backplane connector integrates the mechanical, chemical, and electrical properties
of the device seamlessly . Orientation of differential pairs, spacing of contacts,
and selection of component materials all play key roles in the overall
performance. It is a challenge to find the proper combination of these design
criterion without impacting the signal integrity of the connector. A great
deal of time and effort go into the design and modeling of these types of connectors before the first piece of steel is cut. The exploded isometric drawing
of the connector in Figure 12 is courtesy of ERNI Corporation GmbH.
It is noteworthy to discuss the overall channel performance from transmitter
to receiver within the backplane/daughter card system. A perfectly controlled
impedance environment is the ultimate goal which will result in minimal
reflections, best frequency response, and higher data rates. A typical source
of reflections is thought to be the connectors within a system and this is
certainly possible if the connector is being used at a higher data rate than
recommended. However, the more likely scenario is this: inside the connector
is a well controlled differential transmission line as in Figure 12, but the pcb
to connector interface is poorly designed. This area called the “via field” is
typically the most critical portion of the channel. It can easily make or break
the overall bit error rate performance of the system.
15
Three Configurations Modeled with ADS
Figure 13. Three connector to board interface mechanisms.
The next step is to examine the particular advantages of various connector
to board interface mechanisms including microvias, pressfit, and surface
mount. The diagrams in Figure 13 show the connector to board interface of
a connector receptacle which is soldered to a typical board with 16 layers
and an overall thickness of 4mm. Using this realistic assumption, the following
three typical configurations will be compared. Looking at the table entries
in Figure 11, it can be identified that the selected signal layer is different for
setup 3. The reason for this difference is that a connection to a signal layer
very close to the bottom of a PCB is typically a non-critical case. This is due
to the fact that the remaining stub is very small and the diameter of the via
and the antipad can be optimized to achieve good matching behavior. The
critical case, however, is the wiring from a connector to one of the upper
signal layers in the PCB. Using conventional pressfit or SMT technology
would lead to a via stub with a very large length. The very high capacitive
load of this stub would result in a significant impedance mismatch which
reduces the overall signal quality. This is the reason, why this “worst” case
is not taken into account and the two solutions are compared which are
currently used to overcome this limitation. One of these options is to apply
back drilling to these critical vias while the second one is the implementation
of micro via technology.
16
Characteristic Impedance & Crosstalk
Figure 14. Comparison of the three connector to backplane configurations.
A comparison is done by a simulation in the time domain using a full 3D
high frequency field solver, where a step signal with a rise time of 50 ps
(20 to 80 %) is applied to the device under test. The differential signal is fed
into the remaining part of the connector. Such a waveform is typical for a
10 Gb/s serial transmission. As an outcome of the simulation experiment,
the characteristic impedance Z0 of the system can be evaluated together
with near end crosstalk (NEXT) and far end crosstalk (FEXT), as shown in
Figure 14. Looking at the impedance mismatch, it is evident that the
microvia approach showed the smallest impact on the impedance profile
compared to the other two setups. It can be seen that the crosstalk in the
relatively short vias of the setups with microvias and backdrilled pressfit
connections is much lower compared to the long through hole vias. In both
cases (far end crosstalk, near end crosstalk), the microvia approach showed
a better crosstalk behavior.
17
Differential Eye Diagram Analysis
Figure 15a. Standard via and micro via eye diagrams.
Figure 15b. Very good correlation is shown between PLTS-based eye diagram analysis and
ADS-based eye diagram analysis.
The PLTS eye diagrams in Figure 15a are synthesized from the 4-port
s-parameters obtained for both standard via and microvia structures in
our test vehicles described in Figure 11. PLTS allows multiple views to be
compared side by side for ease of use. The eye diagram is synthesized by
extracting the impulse response of the backplane from it’s s-parameters,
then convolving this impulse response with a user programmable arbitrary
binary sequence. It is useful to have standard PRBS patterns available for
this convolution operation, so PLTS allows a standard PRBS pattern to be
used up to a length of 215-1. This method of creating eye diagrams correlates
well with the standard method of compliance testing with a pattern generator
and a sampling scope with standard masks. As can be seen in Figures 15a
and b, the eye diagrams for the microvia are clearly more open than the
standard via, even at 20 Gb/s.
Eye diagram analysis is very useful when characterizing a device for a
specific digital protocol. For example, we know that Advanced TCA backplanes need to work well at 10 Gb/s. If the eye is sufficiently open and data
transitions don’t encroach into a standardized mask, then we say we have
compliance to that standard. Another useful way to use eye diagram analysis
is determining how much pre-emphasis or equalization is required by the
associated silicon SERDES chipsets to provide compliance. It can be seen
from Figure 15 that the microvia could work at 20 Gb/s with pre-emphasis
or equalization, but most likely the standard via would not.
18
Simulation through Measured/Simulated
S-Parameter Data
Figure 16. ADS generated schematic.
ADS is an environment for combining models and waveforms from many
sources into a single simulation. The ADS schematic shown in Figure 16
describes a simulation environment specifically containing some measurement based models. Measured S-parameter data is used as the model for the
backplane (4-port PLTS), the two daughter cards (4-port PLTS), the coaxial
cable (2-port VNA) and the DCA module as the load (1-port VNA). The card
connectors were modeled as S-parameters using a 3D EM Simulator. The
input waveform is a PRBS source from a BERT tester which was captured
with the DCA-J and imported into ADS.
19
Backplane Model Using ADS
Figure 17. Backplane channel model.
It is necessary to set up the backplane channel model as a subnetwork for
use in a system-level simulation. This is done in ADS by adding ports and
creating a symbol that represents the whole backplane channel as shown
in Figure 17. Note that this complete symbol view is the linear passive interconnect circuit that consists of the daughter cards, backplane connectors,
and the backplane itself.
20
Simulation of Pre-emphasis and Equalization
Through Channel Model
Figure 18. Model with pre-emphasis.
The ADS schematic in Figure 18 shows a Ptolemy co-simulation silicon channel
with transmitter pre-emphasis and receiver equalization. It is common practice
to utilize transmitter pre-emphasis and/or receiver equalization to improve
signal quality of high speed backplanes/channels. This technique will minimize bit errors and open the eye diagram. Through the use of Ptolemy in
ADS, designers can study the effectiveness of different pre-emphasis and
equalization techniques through the channel model. The plots in Figure 18
show the transmitter waveform before and after pre-emphasis. A slider bar is
available in the Ptolemy schematic that controls the amount of pre-emphasis
added to the signal prior to going through the channel model. This gives the
designer an interactive control to see the effect of different amounts of
pre-emphasis and allows optimization of the channel design.
Figure 19. Model with adaptive equalization.
Figure 19 depicts a typical scenario where the receiver applies adaptive
equalization after the signal travels through the channel. The EQ in this case
is provided by a continuously running Least Mean Square (LMS) equalizer.
This could be easily replaced by a Decision Feedback Equalizer (DFE)
design. Such a design may already exist in Matlab or as VHDL code. If so
this could be included directly through a co-simulation.
21
ADS Eye Diagram Front Panel
Figure 20. ADS tool showing rise time.
Eye diagram analysis is currently the most common method of assuring
compliance to digital protocol standards today. For this reason, this figure
of merit is found in both PLTS and ADS tools. The capabilities are complementary and have both been correlated to the industry standard reference
receiver, the Agilent 86100C Digital Communications Analyzer. In brief,
the eye diagram feature in both tools allows a way to view eye diagrams
synthesized from accurate s-parameter measurements. PLTS eye diagrams
are simple and easy to set up quickly, while the ADS tool allows advanced
automatic measurements such as rise time and fall time parameters shown
in Figure 20. In addition, ADS has useful histogram measurements similar to
those found on the DCA-J. The eye diagram measurement algorithms are
aligned with the DCA-J. This strong link to hardware is a critical factor in
avoiding the pitfalls of simulation alone.
22
Summary
Figure 21. Some of the areas that make backplane packaging a challenge.
Original equipment manufacturers (OEMs) have an endless need for increased
system bandwidth, whether for computer servers, mass storage, or internetsupporting equipment such as switches and routers. The backplane packaging
challenge for multigigabit transmissions is to maintain attributes such as
cost and reliability from a copper backplane while accommodating the signal
integrity requirements of passing higher frequency content digital signals.
A proven method of building these larger complex systems is to partition
the design project into smaller tasks including measurements, modeling, and
simulation. Furthermore, a backplane design flow should include both active
and passive component design, co-simulation, pre-build verification, prototype build validation, and final test to compliance standard eye diagrams.
Measurement-based modeling can save time in the design cycle by eliminating the need to build a model of legacy equipment from scratch. Having the
right tools in the signal integrity laboratory will assure success in achieving
the best performing backplane and ultimately the fastest telecommunications
network. It is imperative that the high speed digital designer have a full
understanding of the design tools available today in order to meet these
challenges in the future.
Acknowledgments
Agilent Technologies would like to acknowledge the work of Dr. Thomas
Gneiting of AdMos Technologies in his contribution to this application note.
23
Configuration and Web Resources
PLTS
PLTS version 3.0
•
•
•
•
•
4-port 67 GHz system
One box 4-port 20 GHz (N5230A)
Advanced file export
TDR impedance correction
E-cal support
Website
www.agilent.com/find/plts
Software
• PLTS Analysis Software N1930A
• Advanced Design Software
Hardware
• 4-port 67 GHz VNA-based PLTS
• 4-channel 20 GHz TDR-based PLTS
ADS
ADS 2005A
• ADS Signal Integrity Designer (E9009)
• ADS Signal Integrity Designer Pro (E9010)
Website
• Agilent EEsof EDA home page
http://eesof.tm.agilent.com
• Signal Integrity Applications and Wireline Applications
http://eesof.tm.agilent.com/applications/signal_integrity-b.html
http://eesof.tm.agilent.com/applications/wireline-b.html
• Momentum
http://eesof.tm.agilent.com/products/e8921a-a.html
ADS Training for Signal Integrity
http://www.agilent.com/find/education
• Signal Integrity Class: N3215A Designing for Signal Integrity with ADS
• Ptolemy training: N3206A
Agilent Partners
• Admos for active and passive model extraction services
www.admos.com
Matlab is a U.S. registered trademark of The Math Works, Inc.
www.agilent.com/find/emailupdates
Get the latest information on the products and
applications you select.
For more information about Agilent Technologies test
and measurement products, applications, services, and
for a current sales office listing, visit:
www.agilent.com/find/tmdir
Product specifications and descriptions in this document
subject to change without notice.
Copyright © Agilent Technologies 2005, 2006
Printed in U.S.A. January 18, 2006
5989-4077EN